The EoS MLS Forecaster: Predicting The League … And Week 8 Fixtures

Crystal Ball Predictor

by BILL REESE
Staff Writer

[colored_box color=”red”]Back in the fall, I wrote a column where I used the goal averages of all 10 MLS playoff teams to predict the 2013 MLS Cup champion. This season, I’ll be using those same formulas to forecast which clubs will finish 2014 in a playoff spot, who will raise the Supporters Shield, and who will be playing golf in early November.  Along with the long term view, I will also give weekly favorites for each MLS fixture, starting with … [/colored_box]

MLS Week 8 (April 23–27)

The sample size for my MLS playoff projection model is finally large enough to provide a glimpse of where certain clubs could finish the season if their game-by-game averages continue to hold up. That said, the model can only go on what information it presently has, and as the season rolls along, those forecasts will change dramatically.

For instance, I don’t honestly believe that Real Salt Lake will barnstorm the 2014 regular season and finish undefeated with 84 points. That’s because I’m a human who thinks with reason and rationality. My score simulator, however, only looks at these games based on goals-per-game averages, goals-allowed-averages, and a new quirk I added for 2014—my ”run of form” multiplier. It doesn’t take into account international breaks, summer matinees, friendlies against European superclubs or mid-season DP acquisitions. It operates on the basic assumption that all trends are permanent and irreversible.

Any fan of MLS knows that this league is anything but predictable.

With seven weeks of data to work with, the model projects that the Supporters Shield will go to Real Salt Lake, whose impeccable home form and passable road results bode well for the claret and cobalt. The computer model reflects one of the early trends in the 2014 MLS season—that the West is far better than the East. The model projects that five of the top seven sides in the league table will be Western Conference clubs. The biggest beneficiary of this inter-conference inequality is Sporting KC, who the simulator expects will romp through the east, topping the East by a margin of 13 points. They could theoretically wrap up the conference title (and an automatic berth to the 2015–16 CONCACAF Champions League) before Columbus Day.

The East’s mediocrity is reflected by the amount of ties the simulator has generated. If the computer model has its way, the two play-in game opponents—Philadelphia Union and Houston Dynamo—would both have sub-.500 records and a whopping 16 draws apiece. Hans Backe would be proud!

As the season rolls along, each club’s ratio will alter and many of those tie projections will start turning into wins and losses. The algorithm is designed to be a bit stingy, predicting more wins and losses and fewer ties. The fact that it’s predicted so many draws shows just how evenly mediocre most East clubs are.

As for the hometown New York Red Bulls—the win against Philadelphia at home boosted them from a beta-testing Week 7 projection of a 3-26-5 record to a less infuriating 10-16-8 record. They’re projected to win in this week’s sheet, but posting a clean sheet coupled with a healthy serving of goals will go a long way to pad their home game statistics.

Now, I know predicting the outcome of a 34-game season with 27 games to go is utterly preposterous, but so is signing a 3-year-old soccer toddler to a contract because his father won the Ballon d’Or.

As always, past results do not guarantee future results.

[colored_box color=”red”]

Remember, don’t blame me, blame the math!

Check back every Tuesday for the next EoS Forecast!
Follow Bill’s MLS Prediction generator on Google Docs:
Feel free to leave comments, suggestions, or offer him a job as a data analyst.

[/colored_box]

[highlight color=”eg. yellow, black”]Your Week 8 Predictions![/highlight]

New York > Houston
Seattle < Colorado
Montreal < Philadelphia
DC United < FC Dallas
New England > Sporting KC
Columbus > New York
Real Salt Lake > Vancouver
San Jose = Chivas USA
Houston > Portland

[highlight color=”eg. yellow, black”]Your Week 8 Season-Ending Forecast[/highlight]

Supporters Shield Prediction W L D Pts. Change from Last week
1 REAL SALT LAKE 25 0 9 84 =
2 SPORTING KC 24 3 7 79 =
3 FC DALLAS 25 7 2 77 =
4 COLORADO 21 8 5 68 =
5 COLUMBUS 19 6 9 66 =
6 LA GALAXY 18 6 10 64 =
7 SEATTLE 18 11 5 59 =
8 NEW ENGLAND 14 13 7 49 =
9 PHILADELPHIA 8 10 16 40 =
10 TORONTO FC 8 10 16 40 =
11 VANCOUVER 8 11 15 39 =
12 HOUSTON 11 18 5 38 =
13 NEW YORK 10 16 8 38 =
14 DC UNITED 7 15 12 33 =
15 CHICAGO 5 14 15 30 =
16 SAN JOSE 2 14 18 24 =
17 CHIVAS USA 4 20 10 22 =
18 MONTREAL 2 24 8 14 =
19 PORTLAND 1 24 9 12 =
Eastern Conf. Prediction W L D Pts. Change from Last week
1 SPORTING KC 24 3 7 79 =
2 COLUMBUS 19 6 9 66 =
3 NEW ENGLAND 14 13 7 49 =
4 PHILADELPHIA 8 10 16 40 =
5 TORONTO FC 8 10 16 40 =
6 HOUSTON 11 18 5 38 =
7 NEW YORK 10 16 8 38 =
8 DC UNITED 7 15 12 33 =
9 CHICAGO 5 14 15 30 =
10 MONTREAL 2 24 8 14 =
Western Conf. Prediction W L D Pts. Change from Last week
1 REAL SALT LAKE 25 0 9 84 =
2 FC DALLAS 25 7 2 77 =
3 COLORADO 21 8 5 68 =
4 LA GALAXY 18 6 10 64 =
5 SEATTLE 18 11 5 59 =
6 VANCOUVER 8 11 15 39 =
7 SAN JOSE 2 14 18 24 =
8 CHIVAS USA 4 20 10 22 =
9 PORTLAND 1 24 9 12 =
  • Dave from Dix Hills

    I would die … DIE … If RSL finished undefeated hahahaha!!!!